Traditional Predictor Variables Research Articles

Short‐term forecasting of the US unemployment rate

AbstractThis paper aims to assess whether Google search data are useful when predicting the US unemployment rate among other more traditional predictor variables. A weekly Google index is derived from the keyword “unemployment” and is used in diffusion index variants along with the weekly number of initial claims and monthly estimated latent factors. The unemployment rate forecasts are generated using MIDAS regression models that take into account the actual frequencies of the predictor variables. The forecasts are made in real time, and the forecasts of the best forecasting models exceed, for the most part, the root mean squared forecast error of two benchmarks. However, as the forecasting horizon increases, the forecasting performance of the best diffusion index variants decreases over time, which suggests that the forecasting methods proposed in this paper are most useful in the short term.

Abstract P373: Estimation of Atherosclerotic Cardiovascular Disease Risk in Veterans

There are concerns that 2013 ACC/AHA ASCVD pooled risk prediction models are not accurate predictors in contemporary American patient populations, and that these models do not take into account the highly prevalent use of statin therapy or the effects of statins on vascular disease prevention. Using validated phenotyping algorithms, we evaluated the accuracy of the pooled risk model for the development of Hard ASCVD in 1.42 million Veterans (4.3% women, 16.6% black) with mean age 60 years (range 40-79 years) at baseline VA outpatient clinic evaluations in 2002-2004. There were 93,250 Hard ASCVD events over a mean of 8.4 years of follow up. We estimated Hard ASCVD risk using traditional variables (age, sex, race, total cholesterol, HDL cholesterol, diabetes mellitus, systolic blood pressure, blood pressure treatment, and smoking history) employed in the 2013 ACC/ACC ASCVD risk report. Three analytic approaches were undertaken, using 1) published 2013 ACC/AHA prediction equations, 2) 2013 ACC/AHA variables to fit VA-specific Cox regression models for white men, black men, white women, and black women, and 3) model 2 plus baseline statin use. The Harrell’s c-statistics for the models are shown in the table along with the hazard ratios for statins from model 3. Traditional predictor variables were generally statistically significant in all models, and model 1 over predicted absolute risk of Hard ASCVD in VA by approximately 6%, and statin use was associated with up to an 18% reduction in ASCVD risk. In conclusion, these results demonstrate that the pooled ASCVD prediction model performs relatively well in the VA setting, as shown by relatively good c-statistics. The 2013 ACC/AHA over predicted ASCVD risk and statin therapy at baseline was an important determinant of outcomes.

Prediction of Objectively Measured Physical Activity and Self-Reported Disability Following Lumbar Fusion Surgery

To investigate the predictive value of preoperative fear-avoidance factors (self-efficacy for exercise, pain catastrophizing, kinesiophobia, and depression), walking capacity, and traditional predictor variables for predicting postoperative changes in physical activity level and disability 6 months after lumbar fusion surgery in patients with chronic low back pain (LBP). We prospectively enrolled 118 patients scheduled for lumbar fusion surgery for motion-elicited chronic LBP with degenerative changes in 1-3 segments of the lumbar spine. Associations between the predictors and the dependent variables were investigated with multiple linear regression analysis. Dependent variables were physical activity level as objectively measured with a triaxial accelerometer and disability as measured with the Oswestry Disability Index. Preoperative physical activity level (β=-0.349; P < 0.001) and self-efficacy for exercise (β= 0.176; P= 0.021) were significant predictors of the postoperative change in physical activity. Preoperative disability (β=-0.790; P < 0.001), self-efficacy for exercise (β= 0.152; P= 0.024), and pain catastrophizing (β= 0.383; P= 0.033) were significant predictors for the change in the Oswestry Disability Index. Patients with low levels of preoperative physical activity were more likely to increase their level of physical activity after lumbar fusion surgery, especially when their self-efficacy for exercise was high. However, most of these patients still had low levels of physical activity after surgery, and they may therefore need extra support in increasing their postoperative physical activity levels.

Development of PM2.5 and NO2 models in a LUR framework incorporating satellite remote sensing and air quality model data in Pearl River Delta region, China

High resolution pollution maps are critical to understand the exposure and health effect of local residents to air pollution. Currently, none of the single technologies used to measure or estimate concentrations of pollutants can provide sufficient resolved exposure data. Land use regression (LUR) models were developed to combine ground-based measurements, satellite remote sensing (SRS) and air quality model (AQM), together with geographic and local source related spatial inputs, to generate high resolution pollution maps for both PM2.5 and NO2 in Pearl River Delta (PRD), China. Four sets of LUR models (LUR without SRS or AQM, with SRS only, with AQM only, and with both SRS and AQM), all including local traffic emissions and land use variables, were compared to evaluate the contribution of SRS and AQM data to the performance of LUR models in PRD region. For NO2, the annual model with SRS estimate performed best, explaining 60.5% of the spatial variation. For PM2.5, the annual model with traditional predictor variables without SRS or AQM estimates showed the best performance, explaining 88.4% of the spatial variation. Pollution surfaces at 200 m*200 m resolution were generated according to the best performed models.

Enhanced migratory waterfowl distribution modeling by inclusion of depth to water table data.

In addition to being used as a tool for ecological understanding, management and conservation of migratory waterfowl rely heavily on distribution models; yet these models have poor accuracy when compared to models of other bird groups. The goal of this study is to offer methods to enhance our ability to accurately model the spatial distributions of six migratory waterfowl species. This goal is accomplished by creating models based on species-specific annual cycles and introducing a depth to water table (DWT) data set. The DWT data set, a wetland proxy, is a simulated long-term measure of the point either at or below the surface where climate and geological/topographic water fluxes balance. For species occurrences, the USGS' banding bird data for six relatively common species was used. Distribution models are constructed using Random Forest and MaxEnt. Random Forest classification of habitat and non-habitat provided a measure of DWT variable importance, which indicated that DWT is as important, and often more important, to model accuracy as temperature, precipitation, elevation, and an alternative wetland measure. MaxEnt models that included DWT in addition to traditional predictor variables had a considerable increase in classification accuracy. Also, MaxEnt models created with DWT often had higher accuracy when compared with models created with an alternative measure of wetland habitat. By comparing maps of predicted probability of occurrence and response curves, it is possible to explore how different species respond to water table depth and how a species responds in different seasons. The results of this analysis also illustrate that, as expected, all waterfowl species are tightly affiliated with shallow water table habitat. However, this study illustrates that the intensity of affiliation is not constant between seasons for a species, nor is it consistent between species.

Effectiveness of pupil diameter in a probable‐lie comparison question test for deception

Purpose. There were three objectives of this study: (1) To assess the possibility of using pupil diameter as an index of deception in the context of a comparison question polygraph test. (2) To determine if pupil diameter would make a significant contribution to an optimal multivariate classification equation in combination with the traditional predictor variables used in field polygraph practice. (3) We explored the possibility of replacing one or more of the traditional predictor variables with pupil diameter.Methods. We used a laboratory mock crime experiment with 24 participants, half of whom stole $20 (US) from a secretary's purse. Participants were tested with a comparison question test modelled after standard field practice. Physiological measures were taken with laboratory quality instrumentation. Features were extracted from the physiological measures. Those features were subjected to a number of different statistical analyses.Results. Innocent participants showed larger increases in pupil diameter in response to probable‐lie questions than to relevant questions. Guilty participants did not show differential responding to the question types. The additional of pupil diameter to a multivariate classification model approached, but did not reach significance. Subsequent analyses suggest that pupil diameter might be used to replace the traditional relative blood pressure measure.Conclusions. Pupil diameter was found to be a significant predictor variable for deception. Pupil diameter may be a possible replacement for the traditional relative blood pressure measure. Additional research to explore that possibility would seem to be warranted.

Multi-Institutional Validation of a New Renal Cancer–Specific Survival Nomogram

We tested the hypothesis that the prediction of renal cancer-specific survival can be improved if traditional predictor variables are used within a prognostic nomogram. Two cohorts of patients treated with either radical or partial nephrectomy for renal cortical tumors were used: one (n = 2,530) for nomogram development and for internal validation (200 bootstrap resamples), and a second (n = 1,422) for external validation. Cox proportional hazards regression analyses modeled the 2002 TNM stages, tumor size, Fuhrman grade, histologic subtype, local symptoms, age, and sex. The accuracy of the nomogram was compared with an established staging scheme. Cancer-specific mortality was observed in 598 (23.6%) patients, whereas 200 (7.9%) died as a result of other causes. Follow-up ranged from 0.1 to 286 months (median, 38.8 months). External validation of the nomogram at 1, 2, 5, and 10 years after nephrectomy revealed predictive accuracy of 87.8%, 89.2%, 86.7%, and 88.8%, respectively. Conversely, the alternative staging scheme predicting at 2 and 5 years was less accurate, as evidenced by 86.1% (P = .006) and 83.9% (P = .02) estimates. The new nomogram is more contemporary, provides predictions that reach further in time and, compared with its alternative, which predicts at 2 and 5 years, generates 3.1% and 2.8% more accurate predictions, respectively.

A Century of Stock Market Liquidity and Trading Costs

I assemble an annual time series of bid-ask spreads on Dow Jones stocks from 1900-2000, along with an annual estimate of the weighted-average commission rate for trading NYSE stocks since 1925. Spreads are cyclical, especially during periods of market turmoil. The sum of half-spreads and one-way commissions, multiplied by annual turnover, is an estimate of the annual proportional cost of aggregate equity trading. This cost drives a wedge between aggregate gross equity returns and net equity returns. This wedge can account for only a small part of the observed equity premium, but all else equal the gross equity premium is perhaps 1% lower today than it was early in the 1900's. Finally, I present evidence that the transaction cost measures that also proxy for liquidity - spreads and turnover - predict stock returns one year or more ahead. High spreads predict high stock returns; high turnover predicts low stock returns. These liquidity variables dominate traditional predictor variables, such as the dividend yield. The evidence suggests that time-series variation in aggregate liquidity is an important determinant of conditional expected stock market returns.

The ability of active versus passive coping tasks to predict future blood pressure levels in normotensive men and women

Casual blood pressure (BP) after a 2-year follow-up interval was determined in 40 normotensive men and women (20 Blacks and 20 Whites), who had been initially tested for cardiovascular responses to a variety of active and passive coping tasks, including active speech, passive speech, reaction time, and forehead cold pressor tasks. Stepwise multiple regression analyses were used to identify the best model for predicting follow-up BP. Average systolic blood pressure (SBP) level during cold pressor stress was the single most powerful predictor of casual SBP over 2 years even after controlling for initial resting SBP. Other predictors of follow-up SBP were initial SBP, parental history of hypertension, and heart rate and SBP during passive speech (final model R(2) = .78). For follow-up diastolic blood pressure (DBP), the only significant predictors were initial DBP and male gender. These results contribute to a growing body of literature that suggests that cardiovascular measures observed during stressors have predictive validity above and beyond that of traditional predictor variables.

Survival Among Mexican-Americans, Non-Hispanic Whites, and African-Americans With End-Stage Renal Disease: The Emergence of a Minority Pattern of Increased Incidence and Prolonged Survival

We undertook this study to determine whether there is a significant difference in survival on treatment for end-stage renal disease between Mexican-Americans, non-Hispanic whites, and African-Americans. A database covering the years 1975 to 1986 was obtained from the Texas Kidney Health Program. Eight-eight percent to 90% of patients starting renal replacement therapy in Texas were included in this database. The patients were followed until death, for 3 years after successful transplantation, or until they were lost to follow-up. Life table analysis as well as age-adjusted analysis using the Cox proportional hazards model were performed comparing ethnic/racial groups, disease etiology, and treatment type. In life-table analyses, African-Americans and Mexican-Americans had a survival advantage in most age, disease, and treatment groups. With age adjustment, this survival advantage remained for all etiologies combined, for diabetes and hypertension cases, and for patients receiving hemodialysis in a center. Multivariate analysis revealed a persistent survival advantage for Mexican-Americans independent of traditional predictor variables, such as age, disease etiology, treatment type, or size of the center in which they received treatment. In this same analysis, African-Americans showed an advantage in the older age groups. Both African-Americans and Mexican-Americans on renal replacement therapy have an increased survival advantage compared with non-Hispanic whites. Given the additional burden of increased incidence of end-stage renal disease in these groups, the cost of renal replacement therapy for these minorities is disproportionately high. Further study should be aimed at elucidation of the mechanisms by which minorities achieve their survival advantage.

Knowledge of Senior Center Activities Among the Elderly

This study of the elderly's knowledge of senior center activities shows that the large majority of the aged are unable to identify specific center programs or activities but center users have greater knowledge than nonusers. Traditional predictor variables explain only a small amount of the variation in knowledge. The sources of knowledge cited by the elderly are restricted to newspapers, family and friends, and (in the case of center users) the center publication.

Predicting market behavior: Are psychographics really better?

Research efforts exploring the benefits of life style variables over the more traditional market descriptors have for the most part ignored the question of prediction. If, as has been suggested, psychographics offer the researcher all of the advantages of demographic description combined with the richness and dimensionality of psychological characteristics and depth research, then they should prove to be superior predictors of market behavior. This research examines the predictive ability of life style variables in comparison to more traditional predictor variables in a service market. Using a random example of 1000 applicants for selected bank services, a psychographic profile was constructed through factor analysis. Psychographic, demographic, and combined models were developed with discriminant analysis to predict patterns of bank credit card usage. Results show life style variables to be excellent predictors of user patterns, superior to demographic variables.